[font=Georgia:]Scheme G "first article" DP6037-size case (continued)
The DP6037-size Scheme G enclosure described above has no side walls. I believe that this Scheme G enclosure variant may be the best solution for many users since it assembles securely and comes apart easily.
In case someone does want side walls, however, there's an enclosure variant for that. 1.5 mm acrylic is used for the top, bottom, end, and side windows. There is an opening in the top acrylic window to allow an over-molded mini-USB cable to plug into the board and still hold the USB end window securely in place.
[attachment=2] [attachment=1] [attachment=0] Thanks for your time.
[font=Georgia:]Scheme G "first article" DP6037-size case
The initial "-v1" top and bottom magnet holder prototype parts for Scheme G ordered 06Feb12 from Shapeways arrived back on 23Feb12. This was an unusually long delivery delay even for Shapeways and was probably caused by a catastrophic server failure Shapeways experienced earlier this month.
In fact, these parts took so long to arrive, that I have already updated the top and bottom magnet holder design and ordered new parts as described further on in this forum thread.
These initial "-v1" top and bottom magnet holder prototype parts were used to build a DP6037-size enclosure intended for the Bus Pirate v3.5e design. This enclosure provides 1.5 mm top and bottom acrylic windows [1] so that the circuitry on the top of the board and the pin usage legend on the bottom of the board can be observed. Neodymium magnets in the top and bottom magnet holders sandwich the board through the mounting holes in the corners of the board. For this prototype, the "north" pole protrudes from the bottom magnet holder parts and the "south" pole protrudes from the top magnet holder parts. Although I used cyanoacrylate glue to construct this initial Scheme G enclosure, my intent is to design the various constituent components so that the enclosure that needs no adhesives to hold it together.
[attachment=3] [attachment=2] [attachment=1] [attachment=0] Thanks for your time.
dsm
[1] The top and bottom acrylic windows had already been updated to "-v2" as described elsewhere (link).[/font:]
support via libraries, component warehouses, plug-ins, other programs (e.g. GoogleEarth)
multiple language support
...
In addition to GoogleSketchUp, Autodesk123D, and Inkscape.orgInkscape, I suggest you look into PTCCreo Elements/Direct Modeling Express 4.0 at (link). I have a friend who uses the full version of PTCCreo Elements/Direct Modeling (formerly CoCreate) to build systems with thousands of mechanical parts.
Although creating data to drive a laser-cutter is essentially a 2D operation, using a 3D program allows you to visualize how your parts come together. I've been designing enclosures in 3D using AutodeskInventor Pro 2012 (provided by the TechShop) and then redesigning parts in 2D using CorelCorelDraw (purchased cheaply off eBay) (because the TechShop has a good "printer" driver for this combination of program and laser-cutter). In other words, I personally follow a really poor workflow. Sigh...
Without knowing more about the workflow for the Seeed StudioFusion PCB Service, it is difficult to determine how serious the problem might be. More data is needed. Given that you deal with more Seeed StudioFusion PCB Service boards than most people, maybe you could measure a few boards when you get back home and get a feeling for whether this board size discrepancy issue is real or not. The Seeed StudioFusion PCB Service site tends to list nominal capabilities and not tolerances. There are are a number of sources of error that could creep into their process such as
CAD data round-off errors router path conversion errors board positioning hole diameter tolerance (mechanical) ~ 0.08 mm feature registration (mechanical) ~ 0.09 mm tooling positioning errors router bit wear errors ...
The board edges are a bit rough which makes me suspect router bit wear. The amount of discrepancy we're talking about is only about 10 mils. Removing this amount of material to bring the board into compliance with what is expected using 400 grit sandpaper and a palm sander would be no problem if it were not for the USB connector hanging over the edge (which is a good feature in order to maximize USB engagement margins for typical wall thicknesses). People just need to be aware that there may be a board size tolerance. In most situations, that won't be a problem.
I don't think the source of the problem is anything that Dangerous Prototypes may have done. I do think this issue may just be a side effect of using a very cost effective board fabricator in a very competitive market. The Seeed StudioFusion PCB Service may not even know that they have a problem. Board fabricators are often not very introspective about the limits of their process.
Instead of using a U-shaped main case part and USB end case part to build an enclosure as shown in the previous posting, one could use four individual modular side walls. In both cases, the enclosures would be held together with neodymium magnets. In both cases, the top and bottom laser-cut acrylic windows would be the same size as the main board for easy customization. The top and bottom laser-cut acrylic windows also keep the enclosure from "racking".
Assume for the moment that these modular side walls can be generated parametrically by changing one length related parameter. Dimensions are nominal - actual dimensions include wall thickness adjusted for slot depth and a 0.4 mm board-to-case margin. Cases for all proposed Dangerous Prototypes board sizes could be constructed from the following modular side walls (plus variants with a USB connector opening)...
50 mm modular-side-wall 60 mm modular-side-wall 70 mm modular-side-wall 80 mm modular-side-wall 90 mm modular-side-wall 100 mm modular-side-wall + 31 mm modular-side-wall 37 mm modular-side-wall 43 mm modular-side-wall 49 mm modular-side-wall 56 mm modular-side-wall 62 mm modular-side-wall
[attachment=4] [attachment=3] For example, a DP6037-size case would require the following components...
1x ~ 37 mm modular-usb-side-wall 2x ~ 60 mm modular-side-wall 1x ~ 37 mm modular-side-wall 8x ~ 3.0 mm diameter by 3.0 mm long neodymium magnets 2x ~ 60 mm by 37 mm by 1.5 mm acrylic windows
Ignoring the Shapeways handling fee for a moment,
The laser-sintered nylon components in white for a
DP6037-size case would be roughly $7.62. The laser-sintered nylon components in white for a DP10062-size case would be roughly $12.50.[/list] [attachment=2] [attachment=1] [attachment=0] Thanks for your time.
I like your type 2 stacked style treatment of the interface between the side acrylic and end acrylic. I can see how your type 2 stacked style would be easier to assemble. I also like how your type 2 stacked style avoids the narrow (and fragile) "ears" on your type 1 interlocking style.
I did have one observation worth mentioning derived from the thin laser-cut enclosure examples that I built (link). For thin enclosures with a USB connector, using two tabs on the USB end acrylic may lead to a very thin and fragile acrylic section near the USB connector. For thin enclosures [1], using one wide tab in the end acrylic of smaller cases or a wide center tab in the end acrylic of larger cases may reduce this problem.
For my thin DP6037-size enclosure examples (8.5 mm interior and 11.5 mm exterior), I considered using two 10 mm tabs in the end acrylic...
5.0 mm above-the-board space - (3.95 mm height-of-mini-USB-connector + 0.4 mm margin) =
0.65 mm [/list] But I decided to used one 20 mm wide tab in the end acrylic instead...
0.65 mm (see above) + 1.5 mm tab-length =
2.15 mm [/list] Note that even 2.15 mm is less that the 2x acrylic thickness suggested for the edge-to-slot specification (which should probably also apply to the edge-to-USB-opening dimension).
Thanks for your time.
dsm
[1] Obviously increasing the above-the-board space also works, but since I am interested in relatively thin cases...[/font:]
The DP6037-open-case-dsm-v1 prototype case parts ordered 04Feb2012 from Shapeways arrived back on 14Feb2012. See (link) for more details on this design.
[attachment=0] Case Assembly Procedure
Use an ordinary wooden toothpick to make sure any loose nylon powder left over from the Shapeways SLS 3D printing process is removed from the magnet wells in the U-shaped main case part and the USB end case part. The reason for this step is to make sure that the magnets seat as deeply as possible in the magnet wells. ----------
Press a 3.0 mm diameter by 3.0 mm long neodymium magnet into the left-hand magnet well of the USB end case part (when viewed from the inside with the USB opening toward the top) with the "north" pole of the magnet facing out of the magnet well. Use a small drop of cyanoacrylate (aka super glue) to hold the magnet in position if the magnet well is loose.
Press a 3.0 mm diameter by 3.0 mm long neodymium magnet into the right-hand magnet well of the USB end case part (when viewed from the inside with the USB opening toward the top) with the "south" pole of the magnet facing out of the magnet well. Use a small drop of cyanoacrylate (aka super glue) to hold the magnet in position if the magnet well is loose.
Press a 3.0 mm diameter by 3.0 mm long neodymium magnet into the left-hand magnet well of the U-shaped main case part (when viewed from the inside with top side up) with the "north" pole of the magnet facing out of the magnet well. Use a small drop of cyanoacrylate (aka super glue) to hold the magnet in position if the magnet well is loose.
Press a 3.0 mm diameter by 3.0 mm long neodymium magnet into the right-hand magnet well of the U-shaped main case part (when viewed from the inside with top side up) with the "south" pole of the magnet facing out of the magnet well. Use a small drop of cyanoacrylate (aka super glue) to hold the magnet in position if the magnet well is loose. ----------
Slightly chamfer the top and bottom acrylic windows using 400 grit sandpaper. The reason for this step is to remove the slightly raised edge that the that laser-cutting process sometimes leaves.
Place the IO connector opening in the top acrylic window over the IO connector on Bus Pirate v4c board so that the opening is centered on the connector. Slide this combined assembly into the U-shaped main case part. Note that the U-shaped main case part can be slightly bent to avoid board components (because it is made of nylon).
Slide the bottom acrylic window into the bottom slot in the U-shaped main case part.
Attach the USB end case part to the open end of the U-shaped main case part using the neodymium magnets.
Comments
The orientation of the magnets is intended to make sure that (a) the U-shaped main case part and USB end case part can't be assembled wrong and (b) to allow case parts from different production runs to be assembled correctly.
The neodymium magnets are sufficiently strong that you can determine the "north" [1] and "south" [2] poles of the magnets by just letting the individual magnets roll around on a smooth table.
The prototype DP6037-sized Bus Pirate boards [3] were larger than expected. The prototype Bus Pirate v3.5e board was actually 60.2 mm by 37.3 mm. The prototype Bus Pirate v4c board was actually 60.15 mm by 37.25 mm.
These dimensions are toward the upper end of the range that this case was designed to accommodate. The result is a small gap at the corner that disappears if the prototype Bus Pirate board is removed and top and bottom acrylic windows are still present.
Is this a systematic problem with Seeed StudioFusion PCB Service boards?
[attachment=1] [attachment=2] Thanks for your time.
dsm
[1] "north" pole refers to the end of the magnet that is attracted to the earth's north magnetic pole. [2] "south" pole refers to the end of the magnet that is attracted to the earth's south magnetic pole. [3] The the DP6037-size Bus Pirate boards shown in the images are prototypes since some components on the boards violate the board edge component keepout specification.[/font:]
I made my first order from McMaster-Carr (link) on Monday and the complete order arrived the next day. I was extremely impressed. The McMaster-Carr site is easy to navigate, their range of products is huge, their prices are more or less reasonable, and their ordering process is easy.
[font=Georgia:]sample DP6037-size Bus Pirate boards
ian,
The following bare prototype DP6037-size Bus Pirate boards arrived yesterday (14Feb2012).
Bus Pirate v3.5e Bus Pirate v4c [/list] Thank you very much. These boards will be quite helpful for developing enclosures. I plan to populate only the mini-USB connector and the shrouded header at this point since these boards still have the board edge component keepout violations we've discussed (link).
Have you decided about whether the Bus Pirate v3.5e and Bus Pirate v4c layouts need to be updated as described above? The reason I ask is that some the board edge alternate SOB case designs (which might be the best choice for Bus Pirate v4c) are only possible if the "board edge component keepout" areas are clear.
[font=Georgia:]mortise and tenon joints in acrylic
tayken,
[quote author="dsm"]Adding a slot to longer bottom tabs (secured with another piece of acrylic that spans the box) might help. [/quote] The following acrylic box design from the TechShop examples cabinet illustrates the box closure scheme I was describing above. Think of it as multiple mortise and tenon joints in acrylic. Since acrylic (at the right scale) is slightly flexible, the piece of acrylic that spans the box is slightly flexed to lock it into position.
Sorry. I didn't express myself clearly. I was just agreeing with your observation that the simple tabs and slots approach to these cases with sides was going to have some assembly problems. I didn't mean to imply that there were no possible solutions.
Some possible approaches to make the simple tabs and slots scheme more viable include...
Adjusting the slot width (which you mentioned) might help. Just making the tabs longer might help. Adding one-way barbs to longer bottom tabs might help. Adding a slot to longer bottom tabs (secured with another piece of acrylic that spans the box) might help.
[quote author="tayken"]The interconnection between slots dictate that they must be assembled first, then top & bottom pieces must be mounted. I have an idea about this...[/quote] Redesigning the various parts to have both tabs and slots (link) would allow you to force the order of assembly. For example, bottom into ends, ends into sides, sides into top.
[quote author="tayken"]Now I can understand what my ME professor was trying to say when he was talking about "Design for ease of assembly".[/quote] The problem with DfEoA in this context is that what we really want is DfEoA & D & A & D & A & D & A... Sigh...
[quote author="tayken"]I still have a few tricks up my sleeve to solve the assembly problems.[/quote] Good luck with your prototyping. I'm looking forward to seeing what you develop. If there is anything I can do to help, please don't hesitate to ask.
The prototype cases I described earlier in this forum thread had the following vertical dimension budget:
1.5 mm top acrylic window + 5.0 mm space-above-board + 1.7 mm nominal board thickness + 1.8 mm space-below-board + 1.5 mm bottom acrylic window = 11.5 mm (plus any screw heads, etc.)
I was wondering how much custom 3D printed spacers in SLS nylon might cost for the following...
4x spacers ~ 5.0 mm OD / 3.2 mm ID / 5.0 mm long (
e.g. space-above-board) 4x spacers ~ 5.0 mm OD / 3.2 mm ID / 1.8 mm long (e.g. space-below-board) [/list] Ignoring the Shapeways handling fee for a moment, the cost for a set of eight spacers in white SLS nylon appears to be $0.48. This includes an assembly scheme to keep the spacers together until the end user cuts them apart. The following image shows four sets of the eight spacers described above.
I think tayken realized the implicit assembly problems of these cases with sides before I did. I just ran some experiments because I am interested in thin cases (which implies 1.5 mm acrylic to me).
There are a couple of edge mounting schemes that might work well with Bus Pirate v4c. I have some parts being fabricated by Shapeways to prototype these schemes which should arrive back next week.
I had some time scheduled on the TechShop laser-cutter for another project and finished early, so I built some 1.5 mm acrylic parts for (right to left)...
Scheme G with sides
Scheme G without sides
Alternate SOB size case design
[attachment=1] [attachment=0] I assembled two Scheme G with sides case examples and have a couple comments...
The only difference between the two examples is the width of the slot.
The inside edge of the slots is nominally fixed to the desired inside case length and width.
These cases are the same overall thickness as the yabp v4fd enclosure design (red case in front), namely... 8.5 mm interior == 5.0 mm space-above-the-board + 1.7 mm nominal board thickness + 1.8 mm space-below-board. 11.5 mm exterior == 1.5 mm acrylic-top-window + 8.5 mm interior + 1.5 mm acrylic-bottom-window.
The case examples clearly illustrate the problem that Bus Pirate v4c layout may have with Scheme G. An edge mounting scheme might be a better choice for the Bus Pirate v4c layout. See the Alternate SOB size case design forum thread for more details.
The ends only have one tab and slot instead of the two that I recommended previously because of how close the USB connector is to the top of this fairly thin case.
Although I was able to assemble the case, this would not be a task I would like to do very often. Even if the case parts were attached together with cyanoacrylate glue, I suspect that cases with tab and slot sides may not be particularly user friendly. I believe tayken made a similar comment (link) regarding his case prototypes.
